Acute Kidney Injury There are many occasions when the etiology of acute kidney injury secondary to intrinsic renal disease is not evident de-spite a carefully performed history and physical examination and the availability of information gained from various labo-ratory studies. A biopsy can be very useful in establishing the diagnosis, determining the approach to management, and de ning the prognosis in this clinical setting. Retrospective studies from several centers have revealed that the diagnosis of acute tubular necrosis (ATN) cannot be established clini-cally76,83–85 in 10% to 25% of patients who present with acute kidney injury. A biopsy in this population can be important because other causes of acute kidney injury are revealed, such as crescentic proliferative glomerulonephritis, intersti-tial nephritis, Wegener granulomatosis, polyarteritis nodosa, multiple myeloma, amyloidosis, endocapillary proliferative glomerulonephritis, cortical necrosis, hemolytic-uremic syndrome (HUS), systemic lupus erythematosus (SLE), and thrombotic thrombocytopenic purpura, to list just a few. These diseases usually require an approach to management that is different than that normally employed in uncompli-cated cases of ATN. Occasionally, a biopsy can provide helpful clinical in-formation in patients who appear to have ATN on clinical grounds at initial presentation, but who do not regain renal function after 2 to 3 weeks of supportive therapy, includingdialysis. Thediagnostic possibilities generally are the same as those listed in the preceding paragraph. A careful evaluationof the clinical situation is deemed prudent before a renal biopsy is initiated because this procedure carries a higher risk in the patient with acute uremia. 86 Nephrotic Syndrome A renal biopsy in the clinical setting of an acute nephrotic syndrome not associated with systemic disease is in uenced greatly by the age of the patient.

It is common practice to treat children initially with high-dose corticosteroids, because most younger children have minimal change nephrotic syndrome (MCNS) on a biopsy. The presence of a selective proteinuria and normal renal function and the absence of hypertension strengthen the clinical diagnosis. In children, a biopsy is usu-ally reserved for patients with no response to corticosteroid therapy or in whom the clinical and laboratory features of the illness at the time of initial presentation are distinctly atypi-cal for MCNS. These features would include hypertension, azotemia in the absence of volume depletion, nonselective proteinuria, a highly active urine sediment including red cell casts, and involvement of other organ systems.Most nephrologists believe that the adult nephrotic pa-tient without signs of systemic disease should undergo a bi-opsy before therapy is initiated because the majority of these patients, including elderly persons, 87 have a renal disease other than MCNS. 10 The most frequent cause of the nephrotic syndrome in adults is idiopathic membranous glomerulo-nephritis10,88; other frequent causes include focal segmental glomerular sclerosis (FSGS), membranoproliferative glomeru-lonephritis (MPGN), proliferative glomerulonephritis, immu-noglobulin A (IgA) nephropathy, and amyloidosis. Because the optimal treatment differs in different conditions, a renal biopsy can provide helpful clinical information. Moreover, fewer than one-third of adult patients have MCNS. Thus, if the physician elects to administer a short course of high-dose corticosteroid therapy equivalent to that employed in pediat-ric patients, approximately two-thirds of the patients would not be expected to respond favorably. Despite suggestions to the contrary, 25 we believe the risks associated with the use of corticosteroids or other immunosuppressive agents, such as azathioprine, chlorambucil, cyclosporine A, mycophenolate mofetil, and cyclophosphamide, in this population are too great to justify their use in the absence of a speci c histologic diagnosis. Isolated Proteinuria Isolated nonnephrotic proteinuria of 1 g or less per 24 hours without hematuria or pyuria in an otherwise asymptomatic patient who does not have diabetes mellitus is a relatively common clinical problem. Often, the proteinuria is rst detected during a routine physical examination required for participation

in school athletics, during a preemployment examination, or at the time of application for life insurance. In young adults, orthostatic proteinuria is commonly identi- ed in this presentation, carries a benign prognosis, and does not require a renal biopsy. 89 Otherwise, unless the patientrequests a kidney biopsy for purposes of reassurance, it is currently our policy to merely monitor the clinical course of such patients at periodic intervals of 6 months to 1 year. There is little evidence to suggest that these patients will progress to renal failure or that they are candidates for any type of speci c medical therapy in the absence of impaired renal function. 90 If there is any evidence during follow-up of functional deterioration or the development of additionalclinical signs or symptoms suggesting the presence of a primary renal disease or kidney involvement secondary to systemic disease, the patient is thoroughly reevaluated and is often advised to undergo a kidney biopsy for diagnosis and possible therapeutic intervention. In asymptomatic patients who do not have diabetes melli-tus and who remain nonnephrotic but persistently excrete more than 1 g of protein per 24 hours, we advise a renal biopsy. It is this group of patients who are more likely to have an underly-ing renal abnormality. Some of the more common diagnostic possibilities include early idiopathic membranous glomerulo-nephritis, FSGS, and IgA nephropathy. Patients with urinary abnormalities such as hyaline and granular casts are even more likely to have an underlying glomerular abnormality. 91 Hematuria with or without Proteinuria Asymptomatic hematuria, especially in children and young adults, is a frequent cause of referral to nephrologists. It is im-portant that causes of hematuria due to neoplasms in either the upper or lower collecting system or due to either cystitis or pyelonephritis be excluded before one considers a renal biopsy. In general, the diagnostic value of a renal biopsy in the setting of idiopathic microscopic hematuria relates directly to the extent of associated clinical and laboratory ndings. For example, in a series of 76 pediatric patients with isolated hematuria, Trachtman et al. 92 found that almost three-quarters of all biopsy specimens obtained in patients who had either a rst-degree relative with hematuria or a history of at least one episode of gross hematuria were abnormal histologically. IgA nephropathy and Alport syndrome were the two most com-mon ndings. Schröder et al. 93 performed renal biopsies in 65 children with isolated hematuria of at least a 1-year duration. Of the group, 95% had histologic abnormalities that included IgA nephropathy (16 patients), Alport syndrome (8 patients), thin glomerular basement membrane (33 patients), and non-speci c mesangial abnormalities (5 patients). In a later report, Topham et al. 94 evaluated 165 children and adults with isolated hematuria using cystourethroscopy and renal biopsy. All had a normal intravenous pyelogram, were normotensive with a normal serum creatinine, and were free of both proteinuria and a urinary tract infection. In this group, 47% had signi cant histologic ndings, including IgA nephropathy in 49 patients, whereas only 5 abnormalities were identi ed on a cystoure-throscopy. Renal biopsy abnormalities were most common among patients under 20 years of age (69%), prompting these investigators to conclude that a renal biopsy should replace a cystoscopy in younger patients as the next step in evaluation if renal imaging yielded normal results. Furthermore, because renal histologic abnormalities are quite frequent in the clinical setting of isolated hematuria, these investigatorsrecommended a kidney biopsy in patients over 45 years of age in whom ndings at renal imaging and cystoscopy are normal. The likelihood of identifying signi cant glomerular pa-thology is considerably higher when hematuria is accom-panied by proteinuria, with or without an abnormal urine sediment that includes red blood cell, granular, hyaline, or white blood cell casts. We believe it is important to establish the histologic diagnosis of the renal lesion in this clinical set-ting; although admittedly, a biopsy is not required to identify the source of hematuria. Primary renal diseases that can be seen include IgA nephropathy, acute or resolving postinfec-tious glomerulonephritis, MPGN, and an occasional exam-ple of interstitial nephritis. Heredofamilial and multisystem diseases that may be seen include Fabry disease, sickle cell trait and disease, polyarteritis nodosa, Wegener granuloma-tosis,

diabetes mellitus, SLE, and Henoch-Schönlein disease. Many of these systemic diseases may be evident on clinical grounds if a careful prebiopsy evaluation is undertaken, as discussed in the next section. Systemic Disease There are many systemic diseases that involve the kidney, although the extent and frequency of involvement varies considerably in different conditions. Patients often undergo a renal biopsy for diagnosis and management on the basis of either the frequency or severity of the renal lesion. These dis-eases include SLE, Henoch-Schönlein purpura, polyarteritis nodosa, Goodpasture syndrome, Wegener granulomatosis, and various gammopathies. In approximately 40% to 50% of all patients with type I insulin-requiring diabetes mellitus and comparable percent-ages with type II adult-onset diabetes mellitus, renal failure develops during the course of the disease. 95,96 The natural history of renal disease in both types of diabetes mellitus has been well studied and is reasonably predictable 96; thus, in most patients, a renal biopsy is seldom indicated for a diagnosis or management. However, a biopsy can be helpful in patients whose course may be complicated by the suddendevelopment of renal failure, proteinuria, or nephrotic syn-drome, or who have serologic evidence of other causes of renal disease. Although nephrotic syndrome is observed in approxi-mately 10% of all patients with diabetes, its sudden ap-pearance, especially in the young diabetic without previous evidence of functional renal impairment, should not be ascribed automatically to diabetic nephropathy. This point is well illustrated by the experience of Urizar et al. 97 who described ve young diabetic patients with nephrotic syn-drome in whom the renal disease was not distinguishable histologically from MCNS. Nephrotic syndrome appeared either simultaneously or shortly after the recognized onset of diabetes in three of the children. Treatment with cortico-steroids in four patients resulted in a prompt response, with loss of edema, cessation of proteinuria, and normalization of all serum abnormalities. No patient had abnormalities suggestive of diabetic nephropathy. Other investigators have reported similar experiences. 98,99Other types of renal disease also can be seen in asso-ciation with diabetes mellitus, often in the clinical setting of the nephrotic syndrome. Couser et al. 100 reported the coexistence of dense deposits within the glomerular and tubular basement membranes, resembling those seen in type 2 MPGN and lesions typical of diabetic nephropathy in a 24-year-old nephrotic man with type I diabetes mellitus. Other examples of well recognized renal diseases that have been reported to occur in patients with diabetes mellitus in either the presence or absence of diabetic nephropathy include acute postinfectious proliferative glomerulonephri-tis,101,102 crescentic proliferative glomerulonephritis, 101 and membranous glomerulonephritis. 103–105The renal biopsy is central to the management of SLE with renal involvement (i.e., lupus nephritis). At present, it is our practice to biopsy all patients who present with clinical evidence of active lupus nephritis unless a medical contraindication exists. Border 106 has suggested that patients with more than six red blood cells (RBCs)/high-power eld, a urine protein excretion greater than 200 mg per 24 hours, or an abnormal serum creatinine value are candidates for a biopsy. There is no other way to establish the type of renal lesion that is present, and the management of lupus nephritis varies considerably depending on the speci c histologic lesion. The value of renal biopsy in predicting a prognosis has been debated. The results of earlier studies suggested that the biopsy classi cation of lupus nephritis was useful in predicting the clinical course 107,108 and this issue was chal-lenged109–111 from a prognostic standpoint but reaf rmed subsequently. 112–116 Correspondingly, we believe it is impor-tant to establish as precise a histologic diagnosis as possible because, in general, patients with diffuse proliferative lupus nephritis with signs of disease activity, such as increased cel-lularity, segmental necrosis, brinoid deposits, and crescents in the glomeruli, have a poorer prognosis than individuals with mesangiopathic, focal proliferative, or membranous lupus nephritis. Controversy also exists concerning the value of renal biopsy in patients with clinically silent lupus nephritis. In 1977, Mahajan et al. 117 described 12 patients with diffuse lupus nephritis but

without clinical or laboratory evidence of renal involvement at the time of renal biopsy. A later re-port, in which 10 of the original 12 patients were followed from 5 to 11 years, revealed deterioration of renal function in 3 years, with one death as the result of renal failure. 118 All patients received prednisone alone or in combination with azathioprine. These investigators concluded that the prog-nosis for the preservation of renal function appeared better in patients with clinically silent diffuse proliferative ne-phropathy as opposed to those with clinically active disease, and recommended a biopsy in patients with SLE even in the absence of overt clinical renal involvement. 118 Woolf et al. 119described eight patients ranging in age from 6 to 26 years, with clinically silent lupus nephritis, who on biopsy had a variety of histologic lesions indicative of active renal involve-ment. Although no consensus exists regarding the use of a renal biopsy in patients with SLE who are without clinical evidence of renal involvement, it is currently our policy to withhold a biopsy in this group of patients. Renal biopsy can often aid the clinician in selecting an appropriate therapy for the treatment of vasculitis when re-nal involvement is present. Polyarteritis nodosa and Wegener granulomatosis require aggressive combination therapy with cyclophosphamide and prednisone. The prognostic value of crescents in antiglomerular basement membrane (GBM) dis-ease and other conditions is discussed in later paragraphs. Other systemic diseases that often exhibit renal involvement and, therefore, can be diagnosed with the aid of a renal bi-opsy when other diagnostic tests have failed or have not been employed include multiple myelomas, kappa light-chain disease,120 amyloidosis, 121 brillary glomerulonephritis, and mixed cryoglobulinemia with renal failure. 85,122 Transplant Kidney Renal biopsies are a valuable diagnostic tool in the manage-ment of the transplant recipient. A biopsy of an allograft represents the major clinical exception to avoidance of a per-cutaneous biopsy of a single functioning kidney. Numerous studies con rm the value and relative safety of a renal biopsyin this setting. 41,42–48,123,124 A biopsy is the most accurate means of determining the presence of lesions, such as cellular or humoral rejection, ATN, drug-induced or viral (especiallyBK virus) interstitial nephritis, hemorrhagic infarction, calci-neurin inhibitor toxicity, and de novo or recurrent glomeru-lonephritis in the allograft. There are several clinical settings in which a biopsy of the allograft is often indicated. These include failure of the graft to function within the initial 7 to 10 days after surgery, a rapid deterioration in function of unknown etiology after the initial good function, an absence of a response to an adequate antirejection therapy within a reasonable period of time, and an unexplained nephrotic syndrome or nephrotic-range proteinuria. A large number of cadaveric kidneys are engrafted, and ischemia-reperfusion injury is a frequent complication. Fail-ure to achieve improved renal function within 7 to 10 days after surgery raises the possibility of a more severe form of renal injury, such as an infarction or a superimposed episode of acute rejection. A biopsy is often invaluable in determin-ing the etiology of the renal failure, in guiding subsequent therapy, and in establishing a prognosis. For example, Kiaer et al. 125 reported a 100% graft loss when infarction, capillary thrombosis, and arterial or arteriolar thrombosis were found either singly or in combination on a biopsy. Thus, the pres-ence of these lesions in the clinical setting of an acute kidney injury would obviate the necessity for the continued use of antirejection therapy. The incidence of acute rejection, characterized by a sudden decrease in renal function, is greatest during the rst 6 months after transplantation. In most instances, the sus-picion of acute rejection can be made on clinical grounds. However, acute rejection often occurs in the absence of clinical features, such as graft tenderness or fever, and a patient believed to have acute rejection may not respond to a reasonable course of antirejection therapy. It may be desired to tailor the antirejection therapy to vascular ver-sus tubulointerstitial rejection or antibody versus cellular rejection. A biopsy can be extremely helpful at this junc-ture in the patient’s therapy. In particular, the presence of peritubular C4d deposition suggests the presence of acute humoral rejection,

whereas its absence is typical in cell-mediated rejection. 126,127 A con rmation of acute humoral rejection involves the demonstration of morphologic evi-dence of acute tissue injury in combination with circulating antibodies to either donor human leukocyte antigen (HLA) or to other antidonor endothelial antigens. 128 Other compli-cations, such as ATN, drug-induced nephrotoxicity, or overt renal infarction may be diagnosed. As noted previously, C4d staining of biopsies has been a valuable tool. C4d is produced by the activation of the classic and lectin complement pathways. Thus, ischemia reperfusion (I/R), necrosis, lupus nephritis, and other con-ditions may exhibit C4d staining and must be considered in a biopsy interpretation. 129–133 Current guidelines recom-mend the exclusion of loci of I/R, necrosis, and brosis when using C4d staining to evaluate for possible humoral rejection. 128,134,13 means of determining the presence of lesions, such as cellular or humoral rejection, ATN, drug-induced or viral (especiallyBK virus) interstitial nephritis, hemorrhagic infarction, calci-neurin inhibitor toxicity, and de novo or recurrent glomeru-lonephritis in the allograft. There are several clinical settings in which a biopsy of the allograft is often indicated. These include failure of the graft to function within the initial 7 to 10 days after surgery, a rapid deterioration in function of unknown etiology after the initial good function, an absence of a response to an adequate antirejection therapy within a reasonable period of time, and an unexplained nephrotic syndrome or nephrotic-range proteinuria. A large number of cadaveric kidneys are engrafted, and ischemia-reperfusion injury is a frequent complication. Fail-ure to achieve improved renal function within 7 to 10 days after surgery raises the possibility of a more severe form of renal injury, such as an infarction or a superimposed episode of acute rejection. A biopsy is often invaluable in determin-ing the etiology of the renal failure, in guiding subsequent therapy, and in establishing a prognosis. For example, Kiaer et al. 125 reported a 100% graft loss when infarction, capillary thrombosis, and arterial or arteriolar thrombosis were found either singly or in combination on a biopsy. Thus, the pres-ence of these lesions in the clinical setting of an acute kidney injury would obviate the necessity for the continued use of antirejection therapy. The incidence of acute rejection, characterized by a sudden decrease in renal function, is greatest during the rst 6 months after transplantation. In most instances, the sus-picion of acute rejection can be made on clinical grounds. However, acute rejection often occurs in the absence of clinical features, such as graft tenderness or fever, and a patient believed to have acute rejection may not respond to a reasonable course of antirejection therapy. It may be desired to tailor the antirejection therapy to vascular ver-sus tubulointerstitial rejection or antibody versus cellular rejection. A biopsy can be extremely helpful at this junc-ture in the patient’s therapy. In particular, the presence of peritubular C4d deposition suggests the presence of acute humoral rejection, whereas its absence is typical in cell-mediated rejection. 126,127 A con rmation of acute humoral rejection involves the demonstration of morphologic evi-dence of acute tissue injury in combination with circulating antibodies to either donor human leukocyte antigen (HLA) or to other antidonor endothelial antigens. 128 Other compli-cations, such as ATN, drug-induced nephrotoxicity, or overt renal infarction may be diagnosed. As noted previously, C4d staining of biopsies has been a valuable tool. C4d is produced by the activation of the classic and lectin complement pathways. Thus, ischemia reperfusion (I/R), necrosis, lupus nephritis, and other con-ditions may exhibit C4d staining and must be considered in a biopsy interpretation. 129–133 Current guidelines recom-mend the exclusion of loci of I/R, necrosis, and brosis when using C4d staining to evaluate for possible humoral rejection. 128,134,13The occurrence of the nephrotic syndrome or nephrotic- range proteinuria in a transplant recipient suggests the possi-bility of either recurrent or de novo glomerulonephritis.136,137Those forms of disease that are most likely to recur in the transplant kidney include MPGN, FSGS, diabetic nephropa-thy, and IgA nephropathy. 136,137 To date, the most common de novo disease reported is membranous glomerulonephri-tis.136 Although some would take exception,

we believe it is worthwhile to establish the lesion that is responsible for proteinuria, especially if the proteinuria is associated with a decrease in renal function. Renal Mass or Neoplasm In addition to a percutaneous kidney biopsy for traditional medical indications, as introduced previously, the past decade has seen renewed interest in percutaneous (core) biopsies for renal masses and other neoplasms. The technique fell into dis-favor in previous decades because of bleeding, false-negative results, and other less common complications. 138 Regardless, there is a driving force for tissue diagnosis because 50% of renal neoplasms are now identi ed as incidental to abnormal imaging for other reasons. A percutaneous renal mass biopsy is often performed to evaluate for a possible lymphoma, a renal abscess, or metastatic disease due to a known extrarenal malignancy. It may also be performed to con rm the diagno-sis of a primary renal neoplasm in a patient with known dis-seminated disease or an unresectable retroperitoneal tumor in whom surgical treatment is contraindicated. 138–144 Although there was initial concern regarding the potential for seeding the biopsy tract with malignant cells, only a total of six cases have been reported, and multiple case series published since 1999 have reported no such events. 145 CONTRAINDICATIONS Both the relative and the absolute contraindications for a renal biopsy vary among nephrologists. However, most agree that the risk of complications increases in the presence of severe uncontrolled hypertension, sepsis, known or sus-pected renal parenchymal infection, a hemorrhagic diathesis,a solitary ectopic or horseshoe kidney (except in the case of a transplanted kidney), or when the patient is unable to cooperate during the procedure. In 1958, Kark et al. 146 published the results from their initial 500 percutaneous renal biopsies and listed 11 contra-indications. These included an uncooperative patient, large cysts, a renal neoplasm, a renal artery aneurysm, marked calci c arteriosclerosis, a hemorrhagic diathesis, a single kidney, a perinephric abscess, hydronephrosis or pyonephro-sis, a terminal state of illness, and a rising blood nonprotein nitrogen level greater than 100 mg per deciliter. Hyperten-sion was viewed as a relative contraindication, depending on the importance of the biopsy and the skill of the operator. Certainly, the presence of a single kidney (except a renal allograft), including a horseshoe kidney, sepsis, or a hemorrhagic diathesis, remain important contraindications to percutaneous renal biopsy. 147 However, in some patients, an open or laparoscopic biopsy may well be justi ed if the clinical situation warrants the risk. This also holds true for the patient with a renal artery aneurysm or calci c arterio-sclerosis and an undiagnosed parenchymal renal disease. Many times, a coagulation disorder can be corrected, thus allowing the biopsy to be performed. In most clinical situ-ations, there is little or no reason to perform a biopsy if the patient has large multiple cysts or a terminal illness. The same is true in the presence of a perinephric abscess, acute pyelonephritis, hydronephrosis, or pyonephrosis. Today, a rising BUN or a BUN greater than 100 mg per deciliter is not considered a contraindication if the rise is sudden or unexplained and quite likely due to an acute and poten-tially reversible process, as discussed earlier in this chapter. The presence of normal-sized or large kidneys increases the likelihood that an acute rather than a chronic form of renal failure is present. When a patient is hypertensive, we delay the biopsy until the blood pressure is brought under adequate control. Thus, the presence of hypertension should be considered, at most, a relative contraindication. It is important that blood pressure control is obtained because it is well documented that hypertensive patients are more prone to bleeding after a percutaneous renal biopsy. 17,28 Diaz-Buxo and Donadio 28 not only found a signi cantly greater incidence of complicationsin hypertensive patients (11.6%) as compared with nor-motensive subjects (7.1%) undergoing percutaneous renal biopsy, but the higher incidence also correlated positively with both the severity and the duration of the hypertension. GROSS INSPECTION AND TISSUE PROCESSING An evaluation of a kidney biopsy includes both a gross and a histologic examination of the specimen. A standard histolog-ic examination includes light microscopy, immunohistology, and transmission electron

microscopy. Other less frequently used techniques include scanning electron microscopy, microbiologic cultures, tissue and cell cultures, quantitative or qualitative chemical analyses, enzyme assays, and mo-lecular pathology. Most of the remainder of this chapter is concerned with the preparation, histologic examination, and actual evaluation of the biopsy specimen for clinicopatho-logic interpretation. Gross Examination The general purpose of a gross examination is to determine adequacy and to divide the specimen into the appropriate portions for subsequent processing. The overall dimensions, color, and consistency should be noted. In particular, the area of viable cortex should be identi ed and delineated from the medulla, which is generally paler. Areas of infarc-tion, other necrosis, or pyogenic in ammation that are often pale and highlighted by a hyperemic border may be evident by a gross examination. Ischemia with re ow also may be hyperemic throughout. In general, if the specimen contains both the cortex and the medulla, the medulla is the deeper tissue as it is removed from the needle, although there are several exceptions. If the needle is thrust deeply into the kidney before the core is taken, the cortex may be missed altogether, or if the direction of the needle is obtuse to the pelvis, the needle may pick up the medulla rst and then the cortex as it passes completely through the medulla. We have observed specimens that contain the cortex, the medulla, and then more of the cortex. Chronic disease will make it more dif cult to delineate the cortex. Allografts have additional notable characteristics. First, in older grafts, a thick rind of brous tissue surrounds the graft. This area may be quite pale and should not be confused with the cortex. In newer grafts, the surface may be deeply colored from a hemorrhage or the presence of granu-lation tissue. Second, the outermost rim of the cortex may be pale from ischemic atrophy or necrosis. The deeper cortex is of interest for the diagnosis of an additional disease. Many laboratories employ a dissecting microscope or hand lens to identify glomeruli in the cortex and to guide division of the specimen at the time of the biopsy.

CLINICOPATHOLOGIC CORRELATIONS Once the light, immunohistologic, and electron microscopic ndings are completed, they should be integrated to derive a histologic diagnosis that is indicative of the disease pro-cess. The histologic diagnosis then is related to the clinical ndings to give a clinicopathologic diagnosis that can be used to plan a course of therapy, establish the prognosis, or both. We do not try to describe all of the many histologic patterns of kidney disease in this section, because they are discussed in considerable detail in other chapters of this book. Instead, we brie y discuss selected examples in which the approach we have outlined is used. Several renal diseases fail to reveal signi cant changes or only nonspeci c changes on a histologic examination. For instance, in MCNS, the ndings are principally the result of proteinuria and include foot-process simpli cation in the glomerulus and evidence of increased protein resorption by the proximal tubule. The only abnormality in benign recur-rent hematuria may be the presence of red blood cells in the tubules. The differential diagnosis should include Alport disease early in its course and thin basement membrane disease. Differentiation requires a thorough electron micro-scopic examination of the specimen and the appropriate clinical studies.

Crescents Crescents may be present in every type of immune-mediated glomerulonephritis; therefore, they are not diagnostic. Most types of glomerulonephritis associated with crescents can be identi ed by their characteristic patterns of immunoglobu-lin localization with immuno uorescence microscopy. These include the more severe forms of IgA nephropathy and Henoch-Schönlein purpura (IgA deposits), lupus nephritis, anti-GBM disease, and MPGN. Serologic studies for antineu-trophil cytoplasmic antibodies (ANCA) have helped clarify the diseases with sparse immune deposits (pauci-immune

glomerulonephritis) and brinoid necrosis or crescents. 208These glomerular changes are indistinguishable from the glomerular involvement observed in the microscopic form of polyarteritis nodosa or Wegener granulomatosis. In the ab-sence of a vasculitis involving the muscular arteries in the biopsy specimen, these conditions can be separated by the presence or absence of other systemic organ involvement. 209 association of crescents and the clinical outcome is typi ed by the ndings in anti-GBM disease. Several authors have noted the generally benign course of anti-GBM disease in those few patients who do not develop crescents over the course of their disease even with minimal therapy. 210–212 A close follow-up is prudent in this group because, rarely, a patient has been shown to progress from noncrescentic to crescentic glomerulo-nephritis.211 A graded response in renal and patient survival is dependent on the percentage of crescents. Five studies in the lit-erature 210,212–215 had comparable results that could be combined to evaluate the relationship between the percentage of crescents in the biopsy and renal survival in a total of 133 patients. If the percentage of crescents was less than 85%, most patients (48 of 61) had independent renal function, that is, did not require renal replacement therapy, at follow-up (78%). If the percentage of crescents was 85% or greater, most patients (61 of 72) pro-gressed to renal failure (85%) (Table 13.1). Similar results are found correlating serum creatinine values at presentation with renal survival in anti-GBM disease (Table 13.1). 210,212,214–216There also is a correlation between serum creatinine and the percentage of crescents at presentation. 212In general, the presence of crescents in glomeruli is asso-ciated with a worse prognosis. Exceptions include poststrep-tococcal glomerulonephritis in children where the crescents may resolve without adverse sequelae. 217,218 ANCA-associated glomerulonephritis does not have as clear an association be-tween crescents and renal survival. 208 Therefore, it is preferable to separate the various causes of crescentic glomerulonephri-tis for the determination of prognosis and treatment.